30 Oct
PPC stands for Pay Per Click which is also known as “paid search,” it is a model that lets marketers place ads on an ad platform, paying the platform each time someone clicks the ad.
PPC marketers are often tempted to follow whatever platforms recommend as best practices, but too often these approaches hurt more than help.
Types of PPC (Pay-Per-Click) ads
There are quite a few different types of PPC pay-per-click ads that we should talk about before getting into our list of PPC best practices:
1. Search Ads
These are probably the ads you’re most familiar with. They show up in the search results as text and look just like any other link in the results pages. They are, however, tagged with Ad to show that they’re paid content.
2. Display Ads
Display ads show up on Google’s partner sites and typically show both images and text.
3. Remarketing Ad
Remarketing ads let you customize your search ad campaigns to target previous visitors to your site.
4. Local Ads
These ads show up in local searches with a city name, zip code, or neighborhood in them.
5. Social Ads
Social ads show up in social media feeds and are tagged with Ad or Sponsored.
6. Google Shopping Ads
These ads show up at the top of the search engine results in a carousel format. They let searchers see products and pricing for similar products from a variety of shops.
7. TrueView Ads on YouTube
TrueView ads appear on YouTube videos in the form of fullscreen ads, small sidebars, unskippable ads, or a combination.
8. Sponsored Promotions in Gmail
These ads show up at the top of a user’s mailbox. They look like a regular email but do include an Ad tag.
PPC best practices that are not for your business
1. Relying on broad match
When you’re first starting out in PPC, simply adding keywords without being aware of match types is easy. In fact, both Google and Microsoft will surface recommendations to add broad keywords in ad groups not utilizing that match type.
Broad match comes with the promise of simplicity and expanded reach. However, too often Google Ads or Microsoft Advertising will match completely unrelated search terms to broad keywords. While phrase match and exact match, in their current state, can still match to a number of close variants, these match types still offer greater control than broad.
With that said, a broad match can have its place in the right situation. Combining RLSA targeting with broad match can be effective, as you already know you’re targeting audiences who have expressed some interest in your brand.
In addition, when you’ve maxed out your reach with phrases and exact keywords and still have the budget to test, broad match can be worth trying for discovering potential new queries. You’ll just want to pay careful attention to search terms and watch for irrelevant words and phrases to exclude as negative keywords.
2. Fully automated bidding
Automated bidding has become an unavoidable part of every PPC advertiser’s toolkit on some level. While ad platforms increase the level of automation and decrease the level of control on the advertiser’s end. Don’t blindly accept the default recommendation to use fully automated bidding across the board.
First of all, automated bidding focused around generating conversions (such as Maximize Conversions and Target CPA) depends upon getting accurate conversion data fed back into the ad platform. You need to ensure that you’ve identified the right goals for your brand, whether those be purchases or leads, and that conversion pixels are properly set up to fire for when those occur.
If there are inaccuracies from past conversion tracking setups in your account, you should avoid bid strategies that optimize for conversion performance, as they won’t be working from correct data.
Wait until you’ve had accurate conversion data firing into your account for a few weeks before testing conversion-based bidding. Next, a bid strategy like Maximize Clicks focuses simply around getting the cheapest clicks, and can be detrimental toward actually getting qualified traffic (which might come from pricier keywords).
We’ve found Maximize Clicks helpful in some specific cases, such as for getting display campaigns off the ground, but would generally discourage against this bid strategy for search campaigns.
3. Auto-accepting recommendations
With this ability, advertisers can allegedly save time by letting Google do one more piece of the heavy lifting. With Google auto-applying recommendations, you could potentially see:
- Hundreds of unreviewed keywords added.
- Broad match keywords added where you didn’t intend to use that match type.
- Targeting expansion added to campaigns where you only intended to reach specific audiences.
- And ads created that you didn’t write.
While there are some recommendations from Google that may be helpful for your account, you’re best off taking the time to review through suggestions before implementing them to save later headaches.
4. Including a specific number of keywords per ad group
This is the most frequently asked question, “How many keywords should I include per ad group?”. Like so many questions in this industry, there is no one-size-fits-all answer for this question. You want to think about a few factors here:
- Data significance: Grouping multiple similar keywords together in an ad group allows you to more quickly accrue significant enough impressions, clicks, and conversions for the associated ads to make optimization decisions.
- Intent: If keywords imply different intent, such as top-of-funnel research vs. an immediate need to buy, keep them separate to control bidding and messaging.
- Close variants: As close variant matching has been on the rise, it’s less effective to separate keywords with the same meaning but slightly different wording.
5. Using SKAGs
SKAG stands for Single Keyword Ad Group, this strategy was popularized in some circles within the PPC industry several years ago. The concept behind this approach is that keeping one keyword in one ad group allows you maximum control over the ad messaging that’s presented to that keyword, while also helping you achieve a better quality score to minimize CPCs.
Moreover, SKAGs are a less than optimal tactic for similar reasons to those laid out under the previous point. Close variants on exact and phrase match keywords have practically eliminated any effectiveness SKAGs had in the past.
Even with the best monitoring of search terms, you’re unlikely to catch all of the potential variants that Google may associate with a keyword, and keeping SKAG campaigns and ad groups properly controlled with negative keywords is practically impossible.
SKAGs also often do a disservice by chopping up data for ad performance among many ad groups, where users may have effectively had the same intent via slightly different keywords.Especially as Google continues to automate more bidding and ad rotation, building fewer ad groups with more keywords (as long as they are closely related enough) will give the system more data to optimize around.
6. Including search partners
By default, both Google and Microsoft will opt campaigns into appearing on search partners. These are other sites that have partnered with these respective platforms to include their search results.
The search partners can vary widely in performance depending on the account and sometimes the individual campaign, sometimes resulting in unpredictable spikes in traffic, particularly on the Microsoft Advertising end.
If you have a low budget, you may want to just go ahead and exclude search partners to start. If you have a budget to spare, you may want to test search partners, but you should monitor results closely, looking both at overall conversion rate and cost per conversion, as well as the overall quality of leads.
Note that Google only allows you to turn search partners all on or off at once, without the ability to opt into specific partners. Microsoft Advertising also doesn’t allow you to opt into specific partners, but you can review performance for individual sites and add exclusions.
7. Combining search and display
When setting up a search campaign, Google allows you to opt into their display network within the same campaign. While this approach may seem like an easy way to gain additional reach with minimal effort, too often it results in display placements cannibalizing reach on search.
You’re best off keeping search and display campaigns separated. These networks differ inherently, in that search focuses on direct intent to find a product, while display focuses on audience- and topic-based targeting to surface an offer in front of someone who may or may not be immediately looking for a product.
8. Diversifying into every channel
PPC has come a long way from the basic search and display advertising of the past. Increasingly, online advertisers are spreading spends across the gamut of search, display, social, and native advertising platforms.
While many brands may find a valuable return from launching into a new channel, such as being able to target related Facebook interests on top of those directly searching for a product, you shouldn’t immediately jump to split your spend among several channels without thinking through the “why” first.
Are you willing to invest extra budget into testing a new channel, or would you be pulling budget from an existing channel that’s working well? Are you willing to invest enough budget and time to accurately assess performance on a new channel?
While the ideal amount for a test will vary based on the audience, industry, and ROI goals, you’re not likely to achieve significant enough data with a $500 test. In addition, not every channel makes sense for every brand. For instance, Snapchat probably isn’t the best ad channel for targeting enterprise B2B decision-makers, while LinkedIn likely isn’t the best for selling sports apparel.
Conclusion
Don’t follow the above mentioned recommendations just because a platform suggested you. Whereas, there may be cases where implementing one or more of the above practices may actually help your campaigns.
If you do plan to test a recommendation or “best practice,” make sure to have a clear plan in place for testing, establishing a baseline for results, and evaluating performance to determine whether the practice was worthwhile.